Advertisement

Rifaximin-α reduces gut-derived inflammation and mucin degradation in cirrhosis and encephalopathy: RIFSYS randomised controlled trial

  • Author Footnotes
    † These authors contributed equally.
    Vishal C. Patel
    Footnotes
    † These authors contributed equally.
    Affiliations
    Institute of Liver Studies, King’s College Hospital NHS Foundation Trust, Denmark Hill, London, SE5 9RS, UK

    Institute of Liver Studies, School of Immunology and Microbial Sciences, Faculty of Life Sciences and Medicine, King’s College London, 125 Coldharbour Lane, London SE5 9NU, UK

    The Roger Williams Institute of Hepatology (Foundation for Liver Research), 111 Coldharbour Lane, London, SE5 9NT, UK
    Search for articles by this author
  • Author Footnotes
    † These authors contributed equally.
    Sunjae Lee
    Footnotes
    † These authors contributed equally.
    Affiliations
    Centre for Host-Microbiome Interactions, Dental Institute, King’s College London, UK

    Science for Life Laboratory, KTH - Royal Institute of Technology, 171 21, Stockholm, Sweden

    School of Life Sciences, Gwangju Institute of Science and Technology, Gwangju, 61005, Republic of Korea
    Search for articles by this author
  • Mark J.W. McPhail
    Affiliations
    Institute of Liver Studies, King’s College Hospital NHS Foundation Trust, Denmark Hill, London, SE5 9RS, UK

    Institute of Liver Studies, School of Immunology and Microbial Sciences, Faculty of Life Sciences and Medicine, King’s College London, 125 Coldharbour Lane, London SE5 9NU, UK

    Imperial College London, Biomolecular Medicine, Division of Computational and Systems Medicine, Department of Surgery and Cancer, London, UK
    Search for articles by this author
  • Kevin Da Silva
    Affiliations
    University Paris-Saclay, INRAE, MetaGenoPolis, Jouy-en-Josas, 78350, France
    Search for articles by this author
  • Susie Guilly
    Affiliations
    University Paris-Saclay, INRAE, MetaGenoPolis, Jouy-en-Josas, 78350, France
    Search for articles by this author
  • Ane Zamalloa
    Affiliations
    Institute of Liver Studies, King’s College Hospital NHS Foundation Trust, Denmark Hill, London, SE5 9RS, UK
    Search for articles by this author
  • Elizabeth Witherden
    Affiliations
    Centre for Host-Microbiome Interactions, Dental Institute, King’s College London, UK
    Search for articles by this author
  • Sidsel Støy
    Affiliations
    Aarhus University Hospital, Department of Hepatology and Gastroenterology, Aarhus, Denmark
    Search for articles by this author
  • Godhev Kumar Manakkat Vijay
    Affiliations
    Institute of Liver Studies, School of Immunology and Microbial Sciences, Faculty of Life Sciences and Medicine, King’s College London, 125 Coldharbour Lane, London SE5 9NU, UK
    Search for articles by this author
  • Nicolas Pons
    Affiliations
    University Paris-Saclay, INRAE, MetaGenoPolis, Jouy-en-Josas, 78350, France
    Search for articles by this author
  • Nathalie Galleron
    Affiliations
    University Paris-Saclay, INRAE, MetaGenoPolis, Jouy-en-Josas, 78350, France
    Search for articles by this author
  • Xaiohong Huang
    Affiliations
    Institute of Liver Studies, School of Immunology and Microbial Sciences, Faculty of Life Sciences and Medicine, King’s College London, 125 Coldharbour Lane, London SE5 9NU, UK
    Search for articles by this author
  • Selin Gencer
    Affiliations
    Imperial College London, Biomolecular Medicine, Division of Computational and Systems Medicine, Department of Surgery and Cancer, London, UK
    Search for articles by this author
  • Muireann Coen
    Affiliations
    Imperial College London, Biomolecular Medicine, Division of Computational and Systems Medicine, Department of Surgery and Cancer, London, UK
    Search for articles by this author
  • Thomas Henry Tranah
    Affiliations
    Institute of Liver Studies, King’s College Hospital NHS Foundation Trust, Denmark Hill, London, SE5 9RS, UK

    Institute of Liver Studies, School of Immunology and Microbial Sciences, Faculty of Life Sciences and Medicine, King’s College London, 125 Coldharbour Lane, London SE5 9NU, UK
    Search for articles by this author
  • Julia Alexis Wendon
    Affiliations
    Institute of Liver Studies, King’s College Hospital NHS Foundation Trust, Denmark Hill, London, SE5 9RS, UK

    Institute of Liver Studies, School of Immunology and Microbial Sciences, Faculty of Life Sciences and Medicine, King’s College London, 125 Coldharbour Lane, London SE5 9NU, UK
    Search for articles by this author
  • Kenneth D. Bruce
    Affiliations
    King’s College London, Institute of Pharmaceutical Science, 5th Floor Franklin-Wilkins Building, London, UK
    Search for articles by this author
  • Emmanuelle Le Chatelier
    Affiliations
    University Paris-Saclay, INRAE, MetaGenoPolis, Jouy-en-Josas, 78350, France
    Search for articles by this author
  • Stanislav Dusko Ehrlich
    Affiliations
    University Paris-Saclay, INRAE, MetaGenoPolis, Jouy-en-Josas, 78350, France
    Search for articles by this author
  • Lindsey Ann Edwards
    Affiliations
    Institute of Liver Studies, School of Immunology and Microbial Sciences, Faculty of Life Sciences and Medicine, King’s College London, 125 Coldharbour Lane, London SE5 9NU, UK
    Search for articles by this author
  • Saeed Shoaie
    Affiliations
    Centre for Host-Microbiome Interactions, Dental Institute, King’s College London, UK

    Science for Life Laboratory, KTH - Royal Institute of Technology, 171 21, Stockholm, Sweden
    Search for articles by this author
  • Debbie Lindsay Shawcross
    Correspondence
    Corresponding author. Address: Institute of Liver Studies, School of Immunology and Microbial Sciences, Faculty of Life Sciences and Medicine, King’s College London, 125 Coldharbour Lane, London SE5 9NU, UK; Tel.: +44 (0)20 3299 3713.
    Affiliations
    Institute of Liver Studies, King’s College Hospital NHS Foundation Trust, Denmark Hill, London, SE5 9RS, UK

    Institute of Liver Studies, School of Immunology and Microbial Sciences, Faculty of Life Sciences and Medicine, King’s College London, 125 Coldharbour Lane, London SE5 9NU, UK
    Search for articles by this author
  • Author Footnotes
    † These authors contributed equally.
Open AccessPublished:September 24, 2021DOI:https://doi.org/10.1016/j.jhep.2021.09.010

      Highlights

      • Rifaximin reduced gut-derived systemic inflammation by suppressing oralisation of the gut microbiome.
      • Rifaximin suppressed mucin-degrading species rich in sialidase, e.g. Streptococcus and Veillonella spp.
      • Rifaximin promotes an intestinal environment augmenting responses to pathobionts and promoting gut barrier repair.
      • Patients treated with rifaximin were less likely to develop infections.

      Background & Aims

      Rifaximin-α is efficacious for the prevention of recurrent hepatic encephalopathy (HE), but its mechanism of action remains unclear. We postulated that rifaximin-α reduces gut microbiota-derived endotoxemia and systemic inflammation, a known driver of HE.

      Methods

      In a placebo-controlled, double-blind, mechanistic study, 38 patients with cirrhosis and HE were randomised 1:1 to receive either rifaximin-α (550 mg BID) or placebo for 90 days. Primary outcome: 50% reduction in neutrophil oxidative burst (OB) at 30 days. Secondary outcomes: changes in psychometric hepatic encephalopathy score (PHES) and neurocognitive functioning, shotgun metagenomic sequencing of saliva and faeces, plasma and faecal metabolic profiling, whole blood bacterial DNA quantification, neutrophil toll-like receptor (TLR)-2/4/9 expression and plasma/faecal cytokine analysis.

      Results

      Patients were well-matched: median MELD (11 rifaximin-α vs. 10 placebo). Rifaximin-α did not lead to a 50% reduction in spontaneous neutrophil OB at 30 days compared to baseline (p = 0.48). However, HE grade normalised (p = 0.014) and PHES improved (p = 0.009) after 30 days on rifaximin-α. Rifaximin-α reduced circulating neutrophil TLR-4 expression on day 30 (p = 0.021) and plasma tumour necrosis factor-α (TNF-α) (p <0.001). Rifaximin-α suppressed oralisation of the gut, reducing levels of mucin-degrading sialidase-rich species, Streptococcus spp, Veillonella atypica and parvula, Akkermansia and Hungatella. Rifaximin-α promoted a TNF-α- and interleukin-17E-enriched intestinal microenvironment, augmenting antibacterial responses to invading pathobionts and promoting gut barrier repair. Those on rifaximin-α were less likely to develop infection (odds ratio 0.21; 95% CI 0.05-0.96).

      Conclusion

      Rifaximin-α led to resolution of overt and covert HE, reduced the likelihood of infection, reduced oralisation of the gut and attenuated systemic inflammation. Rifaximin-α plays a role in gut barrier repair, which could be the mechanism by which it ameliorates bacterial translocation and systemic endotoxemia in cirrhosis.

      Clinical Trial Number

      ClinicalTrials.gov NCT02019784.

      Lay summary

      In this clinical trial, we examined the underlying mechanism of action of an antibiotic called rifaximin-α which has been shown to be an effective treatment for a complication of chronic liver disease which effects the brain (termed encephalopathy). We show that rifaximin-α suppresses gut bacteria that translocate from the mouth to the intestine and cause the intestinal wall to become leaky by breaking down the protective mucus barrier. This suppression resolves encephalopathy and reduces inflammation in the blood, preventing the development of infection.

      Graphical abstract

      Keywords

      Introduction

      Advanced cirrhosis brings with it a plethora of complications including hepatic encephalopathy (HE), variceal bleeding, ascites and a propensity to develop infections, which can lead to multiorgan failure. The development of HE in both its covert
      • Patidar K.R.
      • Thacker L.R.
      • Wade J.B.
      • Sterling R.K.
      • Sanyal A.J.
      • Siddiqui M.S.
      • et al.
      Covert hepatic encephalopathy is independently associated with poor survival and increased risk of hospitalization.
      and overt forms
      • Cordoba J.
      • Ventura-Cots M.
      • Simon-Talero M.
      • Amoros A.
      • Pavesi M.
      • Vilstrup H.
      • et al.
      Characteristics, risk factors, and mortality of cirrhotic patients hospitalized for hepatic encephalopathy with and without acute-on-chronic liver failure.
      confers a poor prognosis.
      The gut microbiome has prime importance in the pathogenesis of cirrhosis, with the evolution from a healthy gut microbiome, to one characterised by dysregulated gut microbial activity or ‘dysbiosis’ associated with decompensation of cirrhosis.
      • Woodhouse C.A.
      • Patel V.C.
      • Singanayagam A.
      • Shawcross D.L.
      The gut microbiome as a therapeutic target in the pathogenesis and treatment of chronic liver disease.
      Dysbiosis is greater in patients with cirrhosis who develop complications correlating with plasma endotoxin levels and 30-day mortality.
      • Bajaj J.S.
      • Heuman D.M.
      • Sanyal A.J.
      • Hylemon P.B.
      • Sterling R.K.
      • Stravitz R.T.
      • et al.
      Modulation of the metabiome by rifaximin in patients with cirrhosis and minimal hepatic encephalopathy.
      In cirrhosis there is an imbalance between healthy and pathogenic gut bacteria with skewed microbiota populations in favour of increased numbers of pro-inflammatory and ammonia-producing species including Enterobacteriaceae, Firmicutes, Archaea and Prevotella.
      • Bajaj J.S.
      • Ridlon J.M.
      • Hylemon P.B.
      • Thacker L.R.
      • Heuman D.M.
      • Smith S.
      • et al.
      Linkage of gut microbiome with cognition in hepatic encephalopathy.
      Bacterial translocation (BT) is a significant driver of cirrhosis-associated immune dysfunction (CAID), although the mechanisms by which intestinal dysbiosis drives immune cell dysfunction remain unknown.
      • Albillos A.
      • Lario M.
      • Alvarez-Mon M.
      Cirrhosis-associated immune dysfunction: distinctive features and clinical relevance.
      ,
      • Bellot P.
      • Garcia-Pagan J.C.
      • Frances R.
      • Abraldes J.G.
      • Navasa M.
      • Perez-Mateo M.
      • et al.
      Bacterial DNA translocation is associated with systemic circulatory abnormalities and intrahepatic endothelial dysfunction in patients with cirrhosis.
      Furthermore, there is growing evidence supporting a pivotal role of dysregulated gut microbiota in HE, as well as gut inflammation and barrier dysfunction in decompensated cirrhosis.
      • Riva A.
      • Gray E.H.
      • Azarian S.
      • Zamalloa A.
      • McPhail M.J.
      • Vincent R.P.
      • et al.
      Faecal cytokine profiling as a marker of intestinal inflammation in acutely decompensated cirrhosis.
      The non-absorbable antibiotic rifaximin-α reduces the risk of recurrence of overt HE and need for hospitalisation.
      • Bass N.M.
      • Mullen K.D.
      • Sanyal A.
      • Poordad F.
      • Neff G.
      • Leevy C.B.
      • et al.
      Rifaximin treatment in hepatic encephalopathy.
      Treatment with rifaximin-α has been associated with significant reductions in bed days, emergency department attendances and 30-day readmissions.
      • Orr J.G.
      • Currie C.J.
      • Berni E.
      • Goel A.
      • Moriarty K.J.
      • Sinha A.
      • et al.
      The impact on hospital resource utilisation of treatment of hepatic encephalopathy with rifaximin-alpha.
      The specific mechanism of action of rifaximin-α remains to be elucidated; it has been shown to reduce circulating gut-derived endotoxins
      • Bajaj J.S.
      • Heuman D.M.
      • Sanyal A.J.
      • Hylemon P.B.
      • Sterling R.K.
      • Stravitz R.T.
      • et al.
      Modulation of the metabiome by rifaximin in patients with cirrhosis and minimal hepatic encephalopathy.
      but studies of faecal microbiome composition in response to rifaximin-α have fallen short of demonstrating any distinct changes in microbial abundance utilising 16S rRNA gene sequencing.
      • Bajaj J.S.
      • Heuman D.M.
      • Sanyal A.J.
      • Hylemon P.B.
      • Sterling R.K.
      • Stravitz R.T.
      • et al.
      Modulation of the metabiome by rifaximin in patients with cirrhosis and minimal hepatic encephalopathy.
      ,
      • Kimer N.
      • Pedersen J.S.
      • Tavenier J.
      • Christensen J.E.
      • Busk T.M.
      • Hobolth L.
      • et al.
      Rifaximin has minor effects on bacterial composition, inflammation, and bacterial translocation in cirrhosis: a randomized trial.
      ,
      • Trebicka J.
      • Bork P.
      • Krag A.
      • Arumugam M.
      Utilizing the gut microbiome in decompensated cirrhosis and acute-on-chronic liver failure.
      We hypothesised that rifaximin-α reduces gut microbiota-derived systemic inflammation, a known driver of HE and CAID. A single-centre, double-blind, randomised, placebo-controlled mechanistic trial of rifaximin-α was undertaken on 38 patients with cirrhosis and HE over 90 days to delineate whether rifaximin-α ameliorates neutrophil-derived oxidative stress and systemic inflammation (as a primary objective). Secondary objectives were set to assess changes in HE grade and neurocognitive functioning, as well as to evaluate rifaximin’s mechanism of action by undertaking shotgun metagenomic sequencing (MGS) of faecal and saliva samples, in conjunction with plasma and faecal metabolic profiling, whole blood bacterial DNA quantification, neutrophil toll-like receptor (TLR) expression and plasma and faecal cytokine analysis.

      Patients and methods

      The study was designed to be performed on 50 patients with cirrhosis and chronic HE recruited from King’s College Hospital. 1:1 allocation of rifaximin-α (Targaxan 550 mg) to matching placebo was administered twice daily over 90 days between 15/1/2015 and 20/6/2016 with intention-to-treat analysis. A patient was considered to have cirrhosis if they fulfilled 2 of 3 diagnostic criteria: (i) biochemistry consistent with cirrhosis, (ii) radiology consistent with cirrhosis/portal hypertension and/or (iii) liver histology. The diagnosis of chronic HE was based on the presence of (i) persistent overt HE (≥grade 1) or (ii) ≥2 episodes of overt HE in the previous 6 months.
      Exclusion criteria: age <18 or >75 years, disseminated malignancy (an isolated hepatocellular carcinoma <50 mm was not an exclusion), coeliac or inflammatory bowel disease, intestinal failure, intestinal obstruction and/or previous bowel resection, human immunodeficiency virus infection and chronic granulomatous disease, anti-inflammatory or immunomodulatory drug use, exposure to rifaximin-α in the previous 12 weeks, patients receiving concomitant oral or parenteral antibiotic therapy, known hypersensitivity to rifaximin-α or rifamycin-derivatives, infection with Clostridium difficile or faecal testing positive for Clostridium difficile toxin in the previous 3-months, and pregnancy or breastfeeding women.
      Patient demographics, clinical details (including West Haven HE grade
      • Conn H.O.
      • Leevy C.M.
      • Vlahcevic Z.R.
      • Rodgers J.B.
      • Maddrey W.C.
      • Seeff L.
      • et al.
      Comparison of lactulose and neomycin in the treatment of chronic portal-systemic encephalopathy. A double blind controlled trial.
      ), biochemistry (including venous ammonia) and neutrophil function were assessed at baseline and after 30 and 90 days of rifaximin-α/placebo treatment. Clinically relevant outcomes including overt HE, neurocognitive function by psychometric hepatic encephalopathy score (PHES),
      • Weissenborn K.
      • Ennen J.C.
      • Schomerus H.
      • Ruckert N.
      • Hecker H.
      Neuropsychological characterization of hepatic encephalopathy.
      health-related quality of life (HRQoL), organ failure, infection and mortality were recorded for 90 days.

       Primary endpoint

      A 50% reduction in spontaneous neutrophil production of reactive oxygen species (ROS) 30-days following the start of therapy.

       Secondary endpoints

      Clinical secondary endpoints included HE grade,
      • Conn H.O.
      • Leevy C.M.
      • Vlahcevic Z.R.
      • Rodgers J.B.
      • Maddrey W.C.
      • Seeff L.
      • et al.
      Comparison of lactulose and neomycin in the treatment of chronic portal-systemic encephalopathy. A double blind controlled trial.
      PHES,
      • Weissenborn K.
      • Ennen J.C.
      • Schomerus H.
      • Ruckert N.
      • Hecker H.
      Neuropsychological characterization of hepatic encephalopathy.
      HRQoL
      • Dolan P.
      Modeling valuations for EuroQol health states.
      and incidence of infection and organ failure at 30 and 90 days. Mechanistic endpoints included assessment of systemic inflammation with analyses, at 30 and 90 days, of salivary/faecal microbiome, faecal calprotectin, whole blood bacterial DNA, plasma and faecal metabolome, and neutrophil phenotype and function including circulating TLR-4 expression.

       Ethics and trial registration

      Ethical approval was obtained from NHS Health Research Authority NRES Committee South Central-Oxford C (Bristol) [REC reference:14/SC/0088] and from the Medicines and Healthcare products Regulatory Agency for Clinical Trial Authorisation [EudraCT number: 2013-004708-20; ClinicalTrials.gov NCT02019784]. The trial was conducted in compliance with the principles of the Declaration of Helsinki (1996), principles of Good Clinical Practice, Research Governance Framework and the Medicines for Human Use (Clinical Trial) Regulations. Fully informed consent was obtained from all participants. Some participants eligible for this study were unable to provide informed consent due to cognitive impairment arising from HE and permission from a legal representative was sought.

       Psychometric hepatic encephalopathy score

      A psychometric test battery compromising 5 neurocognitive tests: trail making test A and B, digit symbol substitution test, line tracing test and serial dotting test were performed.
      • Weissenborn K.
      • Ennen J.C.
      • Schomerus H.
      • Ruckert N.
      • Hecker H.
      Neuropsychological characterization of hepatic encephalopathy.

       HRQoL assessment

      The 3-level version of EQ-5D
      • Dolan P.
      Modeling valuations for EuroQol health states.
      consisting of the EQ-5D descriptive system and the EQ visual analogue scale were performed.

       Analysis of neutrophil phenotype and function

      Fluorochrome-conjugated monoclonal antibodies (anti-human CD16, CD11b, IL-8, TLR-2, TLR-4, and TLR-9; BD UK) were used for staining individual patient polymorphonuclear leucocytes from whole blood and analysed by flow cytometry using a FACS Canto II analyser and FACS Diva 6.1.2 software (BD, San Jose, CA). 50,000 granulocytes were gated on forward and side-scatter characteristics and stained with anti-CD16-phycoerythrin-IgG1κ. Fluorochrome mean fluorescence intensity was calculated to detect the receptor binding response and to measure antigen-antibody binding. Neutrophil oxidative burst was quantified using Glycotope Biotechnology Phagoburst™ (BD Biosciences) kits measuring the percentage of phagocytic cells producing ROS at rest.
      • Taylor N.J.
      • Manakkat Vijay G.K.
      • Abeles R.D.
      • Auzinger G.
      • Bernal W.
      • Ma Y.
      • et al.
      The severity of circulating neutrophil dysfunction in patients with cirrhosis is associated with 90-day and 1-year mortality.
      The formation of ROS was detected using the oxidation of dihydrorhodamine-123 to rhodamine-123. Neutrophil phagocytic activity was assessed by neutrophil phagocytosis of opsonized Escherichia coli.
      • Taylor N.J.
      • Manakkat Vijay G.K.
      • Abeles R.D.
      • Auzinger G.
      • Bernal W.
      • Ma Y.
      • et al.
      The severity of circulating neutrophil dysfunction in patients with cirrhosis is associated with 90-day and 1-year mortality.

       Plasma cytokine profiling

      Plasma cytokines were measured using the Meso Scale Discovery (MSD) platform. Samples were run in duplicate on U-PLEX Proinflam Combo 1 (hu) plates, measuring interferon-γ (IFN-γ), interleukin (IL)1-β, IL-2, IL-4, IL-6, IL-8 (CXCL8), IL-10, IL-12 p70, IL-13, and TNF-α.

       Faecal calprotectin

      Faecal calprotectin was measured using the Bühlmann EKCAL2 enzyme-linked immunosorbent assay (EK-CAL, Bühlmann Laboratories, Switzerland).
      • Burri E.
      • Manz M.
      • Rothen C.
      • Rossi L.
      • Beglinger C.
      • Lehmann F.S.
      Monoclonal antibody testing for fecal calprotectin is superior to polyclonal testing of fecal calprotectin and lactoferrin to identify organic intestinal disease in patients with abdominal discomfort.
      The calprotectin cut-off level representing a positive value was 60 μg/g of faeces.

       Faecal cytokines

      Faecal lysates were produced from frozen faecal samples by combined chemical and mechanical homogenisation using an optimised extraction method.
      • Riva A.
      • Gray E.H.
      • Azarian S.
      • Zamalloa A.
      • McPhail M.J.
      • Vincent R.P.
      • et al.
      Faecal cytokine profiling as a marker of intestinal inflammation in acutely decompensated cirrhosis.
      IL-1β, IL-6, IL-10, IL-17A, IL-17E, IL-17F, IL-21, IL-22, IFN-γ and TNF-α were measured in neat faecal lysates using the U-PLEX Th17 Combo 2 (hu) plates and MSD platform.

       Plasma and faecal metabonomic analysis

      Proton nuclear magnetic resonance (NMR) spectroscopy and reversed-phase ultra-performance liquid chromatography coupled to time-of-flight mass spectrometry were undertaken.
      • McPhail M.J.W.
      • Shawcross D.L.
      • Lewis M.R.
      • Coltart I.
      • Want E.J.
      • Antoniades C.G.
      • et al.
      Multivariate metabotyping of plasma predicts survival in patients with decompensated cirrhosis.
      Samples were thawed and prepared for NMR using previously published protocols.
      • Dona A.C.
      • Jimenez B.
      • Schafer H.
      • Humpfer E.
      • Spraul M.
      • Lewis M.R.
      • et al.
      Precision high-throughput proton NMR spectroscopy of human urine, serum, and plasma for large-scale metabolic phenotyping.
      ,
      • Beckonert O.
      • Keun H.C.
      • Ebbels T.M.
      • Bundy J.
      • Holmes E.
      • Lindon J.C.
      • et al.
      Metabolic profiling, metabolomic and metabonomic procedures for NMR spectroscopy of urine, plasma, serum and tissue extracts.

       Whole blood 16S ribosomal DNA quantification

      This was undertaken by quantitative PCR (qPCR) by Vaiomer (Labège, France). DNA was extracted from sterile whole blood. The 16S rDNA present in the samples was measured by qPCR in triplicate and normalised using a plasmid-based standard scale using the workflow described previously.
      • Lelouvier B.
      • Servant F.
      • Paisse S.
      • Brunet A.C.
      • Benyahya S.
      • Serino M.
      • et al.
      Changes in blood microbiota profiles associated with liver fibrosis in obese patients: a pilot analysis.

       Saliva and faecal 16S rDNA and metagenomic species quantification

      16S analysis was undertaken by standard qPCR-based methods. Abundance of MGS, defined as clusters of >500 genes that covary in abundance among individuals, and thus belonging to the same microbial species, was estimated by mapping shotgun sequencing reads onto the genes (performed by R MetaOMineR package). Median signals of the 50 marker genes that represent a robust centroid of gene clusters of MGS were reported (supplementary materials and methods).

       Statistical analysis

      Sample size was determined based on previous in vitro and ex vivo data.
      • Taylor N.J.
      • Manakkat Vijay G.K.
      • Abeles R.D.
      • Auzinger G.
      • Bernal W.
      • Ma Y.
      • et al.
      The severity of circulating neutrophil dysfunction in patients with cirrhosis is associated with 90-day and 1-year mortality.
      Under the assumption of a reduction in spontaneous neutrophil OB from 30% to 15% (constant 60% difference in medians -0.3) and using the Binomial proportions (Exact) method (power 80%; alpha 0.05 [2-tailed t test]), 22 patients were required per study arm.
      Continuous data were tested for normality using the D'Agostino Pearson test. Non-normally distributed data are presented as median (range). A comparison between 2 (or more) groups was done by Student's t test (or Analysis of Variance) and Mann-Whitney U test (or Kruskall Wallis) test for normally and non-normally distributed data, respectively. Comparison between categorical data was done by χ2 test or Fisher's exact test for small sample sizes.
      For continuous data measured over 3 time points, determination of the significance of change was undertaken by repeat measures analysis of variance (RM-ANOVA) with appropriate tests for sphericity. Post hoc tests were used to assess statistical significance between individual time points/groups. Longitudinal ordinal data (e.g. HE grade) was analysed by ordered logistic regression.
      For measures performed at set times using complex laboratory techniques, RM-ANOVA/Student's t test, partial least square discriminant analysis (PLS-DA) and principal component analysis were used. Using ropls R package, metabolomics data were compared.
      Significance was defined at a 95% level and all p values were 2-tailed. Analyses were undertaken utilising IBM SPSS® (version 21).

      Results

       Recruitment

      Eighty-one patients were screened and 38 randomised to rifaximin-α or placebo (Fig. 1) using a web-based block design randomisation system.
      The trial failed to complete recruitment as rifaximin-α was approved in the UK in 2014 for the prevention of recurrent overt HE. Therefore, patients that would have been candidates for participation in the trial were commenced on rifaximin-α as standard of care.

       Patient characteristics

      Patient demographics and baseline characteristics are summarised in Table 1. Patients were well-matched. Fourteen patients were taking lactulose (7/19 [37%] in each arm). There were no significant differences in median MELD (model for end-stage liver disease) score (11 [8-15] rifaximin-α vs. 10 [8-12] placebo), venous ammonia and severity of HE at baseline.
      Table 1Baseline demographic and clinical characteristics by treatment group.
      Rifaximin-α

      n = 19
      Placebo

      n = 19
      p value
      Age58 (52–62)53 (49.5–60.5)0.48
      Male16110.15
      Previous most severe HE grade [0-4]3 (3–3.5)3 (2–3)0.029
      Lactulose771.0
      Proton pump inhibitor15110.3
      Beta blocker1370.10
      Prior TIPS510.18
      Ascites (Yes:No)11:810:90.75
      Previous history of SBP001.0
      Smoking0.91
       Never67
       Stopped88
       Ongoing54
      Alcohol use0.28
       Never31
       Stopped1317
       Ongoing31
      BMI (kg/m2)29.7 (26.3–32.7)26.5 (23.1–29.4)0.068
      Mean arterial pressure (mmHg)87 (78–93)83 (75–86)0.082
      Ascites grade (1-4)1 (1–3)3 (1–3.5)0.25
      Glasgow coma scale (3-15)15 (15–15)15 (15–15)0.29
      Overt HE at day 0 (Yes)14100.31
      White blood cell count [x109/L)6.34 (4.89–7.2)5.44 (4.42–6.25)0.4
      INR1.45 (1.26–1.78)1.37 (1.3–1.67)0.67
      Sodium (mmol/L)139 (137–142)135 (132–137)0.001
      Creatinine (μmol/L)70 (57–87)77 (64–84.5)0.63
      Bilirubin (μmol/L)39 (23–56.5)40 (24–57)0.66
      Albumin (g/L)36 (30–37.5)33 (30–38)0.59
      Venous ammonia (μmol/L)66 (48–78)45.5 (30–64)0.08
      Lactate (mmol/L)1.3 (1.15–1.55)1.7 (1.3–1.95)0.13
      MELD11 (8–15)10 (8–12)0.49
      HE, hepatic encephalopathy; INR, international normalised ratio; MELD, model for end-stage liver disease; SBP, spontaneous bacterial peritonitis; TIPS, transjugular intrahepatic portosystemic shunt. Data are presented as median (range) with comparison between baseline cohorts done by Mann-Whitney U test. Comparison between categorical data was done by the χ2 test. Bold text denotes statistically significant values.

       Primary endpoint

      The trial failed to demonstrate a 50% reduction in spontaneous neutrophil OB at 30 days compared to baseline (p = 0.48) in patients receiving rifaximin-α.

       Rifaximin-α resolved overt HE and improved cognitive function

      No rifaximin-α-treated patients experienced an HE episode compared to 21% (4/19) on placebo. Patients on rifaximin-α normalised their HE grade to zero at 90 days (HE grade 0 [0-1] vs. 0.5 [0-1]; p = 0.014) with an improvement in PHES (p = 0.009) (Table 2). Resolution of HE on rifaximin-α did not translate into an improvement in HRQoL over 90 days.
      Table 2Clinical parameters at baseline, 30 and 90 days post rifaximin-α or placebo.
      VariableBaselineDay 30Day 90Friedman test within group

      p value
      Friedman test within group p value comparing change across 3 time points within group.
      RM-ANOVA within subject effects

      p value
      RM-ANOVA (log transformed for non-parametric data) reflecting within subject effect.
      RM-ANOVA between subject effects

      p value
      RM-ANOVA reflecting between subject comparison.
      HE grade
       Rifaximin-α1 (0–1)0 (0–1)0 (0–0)0.0140.0430.61
       Placebo1 (0–1)0.5 (0–1)0.5 (0–1)0.384
      Trails A (sec)
       Rifaximin-α52 (46–81)48 (36–65)46 (37–54)0.4170.0120.86
       Placebo46 (34–78)46 (37–72)39 (33–61)0.293
      Trails B (sec)
       Rifaximin-α142 (105–161)143 (106–195)144 (94–186)0.880.980.84
       Placebo140 (57–234)135 (73–205)150 (55–194)0.905
      Line tracing (sec)
       Rifaximin-α205 (145–254)185 (111–213)167 (115–270)0.0230.470.56
       Placebo169 (154–255)165 (131–363)135 (120–299)0.496
      Serial dot (sec)
       Rifaximin-α133 (94–178)97 (78–197)102 (74–219)0.2180.940.54
       Placebo101 (83–154)109 (66–189)113 (66–173)0.384
      Digit symbol
       Placebo23 (20–34)24 (19–37)23 (17–38)0.5680.0960.85
       Rifaximin-α21 (16–32)28 (19–36)28 (23–39)0.026
      PHES score
       Rifaximin-α-9 (-13 to -4)-7 (-13 to -3)-6 (-10 to -2)0.0450.0090.617
       Placebo-7 (-13 to -2)-6 (-11to -2)-7 (-12 to -1)0.278
      MELD
       Rifaximin-α11 (8–15)11 (7–14)10 (7–13)0.270.970.99
       Placebo10 (8–12)10 (8–13)11 (8–13)0.076
      White cell count x109/L
       Rifaximin-α6 (3.8–7.6)5.8 (3.3–6.9)6.9 (2.9–6.6)0.320.370.49
       Placebo5 (3.8–5.9)4.3 (3.2–6.3)4.7 (3.8–6.4)0.075
      C-Reactive protein
       Rifaximin-α4.6 (2.8–8.8)5.3 (2.3–12)4.5 (2.4–9.3)0.280.640.96
       Placebo2 (2–9.6)2 (2–4.7)3.1 (2–5.2)0.31
      Neutrophils x109/L
       Rifaximin-α3 (1.8–4.4)2.9 (1.1–3.9)3.1 (1.4–3.9)0.560.570.81
       Placebo2.5 (1.9–4.3)2.5 (1.9–3.8)2.5 (2.1–4.7)0.58
      Creatinine (Μmol/L)
       Rifaximin-α68 (58–78)68 (36–81)69 (55–81)0.990.680.67
       Placebo78 (64–84)86 (64–90)79 (76–92)0.32
      Bilirubin (Μmol/L)
       Rifaximin-α33 (20–53)32 (17–46)29 (24–49)0.550.370.7
       Placebo35 (20–46)32 (24–47)29 (22–47)0.41
      INR
       Rifaximin-α1.4 (1.2–1.8)1.4 (1.2–1.7)1.3 (1.2–1.5)0.0620.490.55
       Placebo1.3 (1.2–1.4)1.4 (1.3–1.5)1.3 (1.2–1.6)0.58
      Venous ammonia (Μmol/L)
       Rifaximin-α62 (49–74)53 (34–72)63 (41–85)0.0230.960.39
       Placebo44 (31–59)58 (42–74)52 (33–71)0.024
      HE, hepatic encephalopathy; INR, international normalised ratio; MELD, model for end-stage liver disease; PHES, psychometric hepatic encephalopathy scoring; RM-ANOVA, repeated measures-ANOVA.
      p <0.05 represents significant difference between groups. Bold text denotes statistically significant values.
      # Friedman test within group p value comparing change across 3 time points within group.
      RM-ANOVA (log transformed for non-parametric data) reflecting within subject effect.
      Δ RM-ANOVA reflecting between subject comparison.

       Rifaximin-α reduced systemic inflammation without changing blood ammonia concentration

      Plasma TNF-α fell significantly at day 30 and 90 (all p <0.001) on rifaximin-α compared to placebo (p <0.001 [Fig. 2A]) with a reduction in IL-10 at day 30 (p = 0.005) which normalised by day 90 (Table 3). Whilst there were no changes in whole blood bacterial DNA levels, there was a significant reduction in circulating neutrophil TLR-4 expression (p = 0.0021) at day 30 in the rifaximin-α-treated patients but not in placebo-treated patients (Fig. 2B). There were no significant changes in circulating neutrophil TLR-2, TLR-9 or IL-8 expression. Those on rifaximin-α were less likely to develop an infection (3 vs. 9); odds ratio for developing an infection on rifaximin-α was 0.21 (95% CI 0.05–0.96) compared to placebo. There were no significant differences in venous ammonia levels between the treatment and placebo arms.
      Figure thumbnail gr2
      Fig. 2Rifaximin-α reduced systemic TNF-α and neutrophil TLR-4 expression.
      (A) Plasma TNF-α fell on rifaximin-α at day 30 and 90 (RM-ANOVA; all p <0.001) compared to placebo. (B) MFI histograms (top panels) and FACS plots (bottom panels) comparing neutrophil TLR-4 expression (MFI) observed at day 30 (blue) compared to baseline (red) in a rifaximin-α-treated compared to a placebo-treated patient. MFI, mean fluorescence intensity; RM-ANOVA, repeated measures-ANOVA; TLR-4, Toll-like receptor-4; TNF-α, tumour necrosis factor-α.
      Table 3Inflammatory indices at baseline, 30 and 90 days post rifaximin-α or placebo.
      VariableBaselineDay 30Day 90Friedman test within group

      p value
      Friedman test within group p value comparing change across 3 time points within group.
      RM-ANOVA within subjects effect

      p value
      RM-ANOVA (log transformed for non-parametric data) reflecting within subject effect.
      RM-ANOVA between subjects effect

      p value
      RM-ANOVA reflecting between subject comparison.
      TNF-α (pg/ml)
       Rifaximin-α4 (3.1–5.3)3.4 (2.8–4.1)3.3 (2.5–3.8)<0.001<0.0010.717
       Placebo4.3 (3–5.9)3.5 (3–5.1)3.7 (3–4.5)0.578
      IL-8 (pg/ml)
       Rifaximin-α34 (28–50)27 (18–68)29 (20–47)0.4090.5470.811
       Placebo38 (23–84)30 (20–114)25 (21–1070.733
      IL-6 (pg/ml)
       Rifaximin-α4.1 (1.8–10.8)3.7 (2.7–4.7)3.8 (2–5.4)0.9350.4120.239
       Placebo9.1 (2.8–19.1)7.1 (2.9–8.3)6.4 (2.3–9.7)0.384
      IL-10 (pg/ml)
       Rifaximin-α0.42 (0.23–0.57)0.23 (0.17–0.19)0.4 (0.19–0.47)0.0050.2160.076
       Placebo0.61 (0.3–1)0.48 (0.21–0.77)0.44 (0.22–0.98)0.274
      IFN-γ (pg/ml)
       Rifaximin-α18 (15–37)19 (12–35)16 (10–37)0.9350.2060.911
       Placebo23 (16–36)24 (16–39)17 (9–104)0.039
      Bacterial DNA (x103)
       Rifaximin-α3.2 (1.7–4.6)3.5 (1.1–4.5)3.3 (1.2–4.3)0.1810.4470.717
       Placebo2.2 (1.6–2.7)2.5 (1.8–3.7)2.9 (2.1–3.9)0.076
      Faecal calprotectin (μg/g of faeces)
       Rifaximin-α105 (55–155)44 (14–129)71 (14–166)0.1760.6580.58
       Placebo116 (48–211)40 (24–149)122 (24–149)0.032
      IFN-γ, interferon-γ; IL-, interleukin-; RM-ANOVA, repeated measures-ANOVA; TNF-α, tumour necrosis factor-α.
      p <0.05 represents significant difference between groups. Bold text denotes statistically significant values.
      # Friedman test within group p value comparing change across 3 time points within group.
      RM-ANOVA (log transformed for non-parametric data) reflecting within subject effect.
      Δ RM-ANOVA reflecting between subject comparison.

       Rifaximin-α led to significant changes in faecal and salivary microbiome whilst preserving beta diversity

      Rifaximin-α reduced species richness compared to placebo in both faeces (Fig. 3A) and saliva (Fig. 3B). Global beta diversity was preserved in the rifaximin-α-treated cohort, but significantly reduced in both faeces (p <0.05 day 90) and saliva (p <0.05 at day 30 and 90) in the placebo-treated cohort (Fig. S1). At the phylum level, rifaximin-α increased faecal Tenericutes and decreased Verrucomicrobia (p <0.05) (Table S1). At the genus level, significant reductions in mucin-degrading genera, such as Veillonella, Akkermansia and Hungatella, were observed in the faecal samples (p <0.05) (Fig. S3 and Table S2). In the saliva, rifaximin-α reduced opportunistic pathogenic genera including Filifactor and Abiotrophia (Table S3). Three distinct genus-based microbial clusters were identified in the faeces (enterotypes): Prevotella, Bacteroides and Firmicutes (Fig. S4). Rifaximin-α enriched the firmicutes enterotype (Fig. 3C). Similarly, 3 distinct microbial clusters were identified in the saliva (oraltype): Prevotella, Neisseria and Lactobacillus (Fig. S4) with rifaximin-α enriching Lactobacillus (Fig. 3D).
      Figure thumbnail gr3
      Fig. 3Rifaximin-α led to significant changes in the microbial community.
      Rifaximin-α reduced species richness in both (A) faecal and (B) salivary microbiome not observed with placebo (Wilcoxon signed rank one-sided tests; ∗p value <0.05). Three distinct microbial clusters in the faecal microbiome were identified enriched with Prevotella, Bacteroides and Firmicutes genera, named enterotype (C). Likewise, 3 distinct microbial clusters were identified in the salivary microbiome enriched with Prevotella, Neisseria and Lactobacillus, named oraltype (D). Fractions of different enterotypes/oraltypes were changed by rifaximin-α, e.g. increasing Firmicutes-type in the faecal and Lactobacillus-type in the salivary microbiome. ET, enterotype; OT, oraltype. ∗ Wilcoxon signed rank one-sided test p value <0.05

       Rifaximin-α suppressed growth of orally originating species in the gut with mucin-degrading capacities

      Rifaximin-α suppressed the growth of orally originating species in the faeces with mucin-degrading capacities and virulence at day 30 and 90, including Veillonella spp and Streptococcus spp as well as Akkermansia and Hungatella (Fig. 4A-C; Fig. S2; Tables S2,4,6). In the saliva, rifaximin-α decreased the opportunistic pathogens Abiotrophia defectiva, Olsenella uli and Filifactor alocis, and significantly increased oral commensal species such as Streptococcus spp (Fig. 4B,D; Tables S5 and S7]. We determined the mucin-degrading capacity of those significantly contrasted species based upon the carbohydrate-active enzyme (CAZyme) annotations of the given species such as sialidase (GH33). The CAZyme families that degrade O-glycans of human mucins are shown in Fig. 4E. Most gut and oral species associated with increased plasma TNF-α and neutrophil TLR-4 expression were enriched with sialidase (GH33) and other mucin-degrading CAZymes (GH2/GH20/GH92/GH130/GH18/GH29 and CBM50). For example, 94% and 81% of the co-abundant gut and oral microbes were enriched with mucin-degrading CAZymes and 19% and 27% were enriched with sialidases, respectively [Fig. S7].
      Figure thumbnail gr4
      Fig. 4Rifaximin-α suppressed growth and mucin-degrading capacity of orally originating species in faeces.
      (A) The number of significantly contrasted species in the faeces/saliva microbiome following rifaximin-α between baseline and days 30/90 are indicated in Venn diagrams. Significantly contrasted species were identified by comparing MGS abundance between baseline and day 30 or 90 by Wilcoxon signed rank tests (p <0.05). Five species in faeces and 21 species in saliva were significantly contrasted on both day 30 and 90. (B) Log2 FC of species abundance following rifaximin-α treatment between baseline and day 30 and 90 (Wilcoxon signed rank test; p <0.05). Significantly contrasted species at both day 30 and 90 are coloured blue (faeces) and orange (saliva). (C) Boxplots for relative abundances (%) of the orally originating species in the faeces i.e. Veillonella atypica and Veillonella parvula in those treated with rifaximin-α vs. placebo. (D) Boxplots for relative abundances (%) of opportunistic pathogens in the saliva i.e. Filifactor alocis and Olsenella uli. (E) The mucin-degrading capacity of significantly contrasted species in faeces and saliva based on CAZyme annotations of given MGS. The CAZyme families that degrade O-glycans of human mucins (black cells of heatmap) represent microbes with mucin-degrading CAZyme classification. CAZyme, carbohydrate-active enzyme; FC, fold change; MGS, metagenomic sequencing.

       Rifaximin-α enhanced faecal cytokines suppressing pathobionts associated with reduced plasma lactate

      PLS-DA revealed differing enrichments of plasma metabolites over time between the rifaximin-α- and placebo-treated cohorts, such as decreased lactate with no substantial changes in acetoacetate, phosphocholine and trimethylamine-N-oxide, which increased over time in the placebo cohort (variable importance of projection >1 and fold change >5%) (Fig. 5A; Fig. S5; Tables S8-11]. No changes in plasma bile acids were seen (data not shown).
      Figure thumbnail gr5
      Fig. 5Plasma metabolome and gut microbiome associations with faecal cytokines.
      (A) Enriched or depleted plasma metabolites at day 30 or 90, compared to baseline. Based on PLS-DA, enriched or depleted metabolites were identified from the 2 study arms. D-lactate, a marker of intestinal barrier damage
      • Riva A.
      • Gray E.H.
      • Azarian S.
      • Zamalloa A.
      • McPhail M.J.
      • Vincent R.P.
      • et al.
      Faecal cytokine profiling as a marker of intestinal inflammation in acutely decompensated cirrhosis.
      was consistently depleted with rifaximin-α, whereas acetoacetate, phosphocholine, and trimethylamine-N-oxide were only enriched with placebo. Log10 FC of significantly altered metabolites (VIP >1) between baseline and day 30 or 90 were coloured on the heatmap from blue (negative value) to red (positive value). (B) Circos plot demonstrating association (Rm2>10%) of species abundances in faeces with faecal cytokines. Weights of the edges from explained variances of fixed effect variables were calculated with linear-mixed effects models by lme4 R package and visualized by the circlize R package. FC, fold-change; IFN-γ, interferon-γ; IL-, interleukin-; PLS-DA, partial least square discriminant analysis; TNF-α, tumour necrosis factor-α; VIP, variable importance of projection.
      Rifaximin-α enhanced day-30 faecal TNF-α (p = 0.0058) and IL-17E (p = 0.011) concentrations and suppressed faecal Veillonella and Streptococcus spp.
      Increased faecal IL-17A (p <0.05), which is important for neutrophil recruitment and augmentation of antibacterial responses to pathogenic bacteria,
      • Edwards L.A.
      • Nistala K.
      • Mills D.C.
      • Stephenson H.N.
      • Zilbauer M.
      • Wren B.W.
      • et al.
      Delineation of the innate and adaptive T-cell immune outcome in the human host in response to Campylobacter jejuni infection.
      was observed on rifaximin-α (Fig. 5B; Table S12).
      Whilst baseline faecal calprotectin was elevated (>60 μg/g) in the majority of patients, consistent with chronic intestinal inflammation, levels did not change on rifaximin-α. No changes were seen in faecal water metabolites on rifaximin-α.

       Adverse events

      Recorded AEs were almost twice as likely in the placebo-treated group (n = 33 placebo vs. n = 17 rifaximin-α). Infection-related AEs were more frequent on placebo. Only 1 serious AE was recorded; small bowel perforation in 1 participant treated with rifaximin-α. This was assessed clinically as a spontaneous event unrelated to the study medication.

      Discussion

      In this double-blind, randomised, placebo-controlled mechanistic trial of rifaximin-α vs. placebo in patients with cirrhosis and HE, rifaximin-α improved HE at 30 days in association with a reduction in biomarkers of gut-derived systemic inflammation, including plasma TNF-α and neutrophil TLR-4 expression. Rifaximin-α suppressed growth of opportunistic orally originating pathogens that were identified in cirrhotic faeces including Veillonella atypica, Veillonella parvula and Streptococcus spp, as well as Akkermansia and Hungatella, all of which are rich in sialidase that degrades O-glycans in the gut mucin barrier. In the saliva, rifaximin-α decreased opportunistic pathogens including Abiotrophia defectiva, Olsenella uli and Filifactor alocis, and led to a significant increase in Lactobacillus and Streptococcus spp associated with oral health. Furthermore, rifaximin-α changed the intestinal microenvironment, resulting in an increase in faecal TNF-α and IL-17E; increased faecal IL-17A being associated with reduced faecal Veillonella and Streptococcus spp.
      Whilst the efficacy of rifaximin-α in reducing the risk of recurrent overt HE is well-established,
      • Bass N.M.
      • Mullen K.D.
      • Sanyal A.
      • Poordad F.
      • Neff G.
      • Leevy C.B.
      • et al.
      Rifaximin treatment in hepatic encephalopathy.
      its mechanism of action remains to be elucidated. Rifaximin-α reduces circulating endotoxin levels
      • Bajaj J.S.
      • Heuman D.M.
      • Sanyal A.J.
      • Hylemon P.B.
      • Sterling R.K.
      • Stravitz R.T.
      • et al.
      Modulation of the metabiome by rifaximin in patients with cirrhosis and minimal hepatic encephalopathy.
      but previous studies have fallen short of demonstrating any distinct changes in microbial abundance by 16S rRNA faecal microbiota profiling.
      • Bajaj J.S.
      • Heuman D.M.
      • Sanyal A.J.
      • Hylemon P.B.
      • Sterling R.K.
      • Stravitz R.T.
      • et al.
      Modulation of the metabiome by rifaximin in patients with cirrhosis and minimal hepatic encephalopathy.
      ,
      • Kimer N.
      • Pedersen J.S.
      • Tavenier J.
      • Christensen J.E.
      • Busk T.M.
      • Hobolth L.
      • et al.
      Rifaximin has minor effects on bacterial composition, inflammation, and bacterial translocation in cirrhosis: a randomized trial.
      One study demonstrated a significant increase in potentially beneficial serum fatty acids and intermediates of carbohydrate metabolism with rifaximin-α.
      • Bajaj J.S.
      • Heuman D.M.
      • Sanyal A.J.
      • Hylemon P.B.
      • Sterling R.K.
      • Stravitz R.T.
      • et al.
      Modulation of the metabiome by rifaximin in patients with cirrhosis and minimal hepatic encephalopathy.
      This is the first study utilising shotgun metagenomic sequencing to explicitly identify changes in the salivary and faecal microbiome in response to rifaximin-α. This was associated with reduced systemic inflammation as evidenced by a reduction in plasma TNF-α and neutrophil TLR-4 expression, surrogate markers for a reduction in circulating endotoxin. Our a priori hypothesis had been that rifaximin-α would reduce neutrophil ROS, as circulating neutrophil dysfunction in cirrhosis has been shown to determine 90-day and 1-year mortality.
      • Taylor N.J.
      • Manakkat Vijay G.K.
      • Abeles R.D.
      • Auzinger G.
      • Bernal W.
      • Ma Y.
      • et al.
      The severity of circulating neutrophil dysfunction in patients with cirrhosis is associated with 90-day and 1-year mortality.
      Whilst not proven, the study was underpowered to reach this endpoint, after rifaximin-α was introduced into national clinical guidelines, making completion of recruitment difficult. However, the improvement in systemic inflammation and neutrophil TLR-4 expression in response to rifaximin-α was evident.
      A change in gut microbiome composition and function impacts on a multitude of vital homeostatic functions including immunomodulation.
      Human Microbiome Project C
      Structure, function and diversity of the healthy human microbiome.
      Patients with cirrhosis have gut dysbiosis with small bowel bacterial overgrowth and translocation of bacteria and their products (such as lipopolysaccharide and bacterial DNA) across a more permeable gut epithelial barrier, exacerbated by underlying portal hypertension and endothelial dysfunction.
      • Bellot P.
      • Garcia-Pagan J.C.
      • Frances R.
      • Abraldes J.G.
      • Navasa M.
      • Perez-Mateo M.
      • et al.
      Bacterial DNA translocation is associated with systemic circulatory abnormalities and intrahepatic endothelial dysfunction in patients with cirrhosis.
      This culminates in systemic inflammation and endotoxemia, inducing CAID through TLR signalling, predisposing to infection, and the development of hepatic decompensation. In this study, patients treated with rifaximin-α experienced fewer infections during the 90-day follow-up period than patients on placebo. This is in keeping with published experience that highlights a potential role for rifaximin-α beyond the prevention of overt HE by augmenting intestinal barrier function and reducing BT and CAID.
      • Salehi S.
      • Tranah T.H.
      • Lim S.
      • Heaton N.
      • Heneghan M.
      • Aluvihare V.
      • et al.
      Rifaximin reduces the incidence of spontaneous bacterial peritonitis, variceal bleeding and all-cause admissions in patients on the liver transplant waiting list.
      ,
      • Menshawy A.
      • Mattar O.
      • Barssoum K.
      • AboEl-Naga A.M.
      • Salim H.M.
      • Mohamed A.M.F.
      • et al.
      Safety and efficacy of rifaximin in prophylaxis of spontaneous bacterial peritonitis: a systematic review and meta-analysis.
      Furthermore, plasma lactate, a metabolic biomarker of poor prognosis associated with non-survival in decompensated cirrhosis
      • McPhail M.J.W.
      • Shawcross D.L.
      • Lewis M.R.
      • Coltart I.
      • Want E.J.
      • Antoniades C.G.
      • et al.
      Multivariate metabotyping of plasma predicts survival in patients with decompensated cirrhosis.
      was reduced by rifaximin-α.
      Gut microbiome pertubations are linked to the pathogenesis of cirrhosis and progression to advanced liver disease.
      • Bajaj J.S.
      • Heuman D.M.
      • Hylemon P.B.
      • Sanyal A.J.
      • White M.B.
      • Monteith P.
      • et al.
      Altered profile of human gut microbiome is associated with cirrhosis and its complications.
      Over 75,000 microbial genes differ between patients with cirrhosis and healthy individuals, with over 50% taxonomically assigned bacterial species of oral origin, suggesting an invasion of the distal gut from the mouth in cirrhosis.
      • Qin N.
      • Yang F.
      • Li A.
      • Prifti E.
      • Chen Y.
      • Shao L.
      • et al.
      Alterations of the human gut microbiome in liver cirrhosis.
      Salivary dysbiosis is also observed in patients with cirrhosis.
      • Bajaj J.S.
      • Betrapally N.S.
      • Hylemon P.B.
      • Heuman D.M.
      • Daita K.
      • White M.B.
      • et al.
      Salivary microbiota reflects changes in gut microbiota in cirrhosis with hepatic encephalopathy.
      Our data confirm salivary dysbiosis – bacteria that are normally resident in the oral cavity translocating to the gut and associated with the generation of a systemic inflammatory milieu. Rifaximin-α suppressed the growth of these orally originating species in the faeces including Veillonella atypica, Veillonella parvula and Streptococcus spp rich in sialidases, enzymes that degrade O-glycans of human mucins. These species are commonly found in dental plaque
      • Chalmers N.I.
      • Palmer Jr., R.J.
      • Cisar J.O.
      • Kolenbrander P.E.
      Characterization of a Streptococcus sp.-Veillonella sp. community micromanipulated from dental plaque.
      being associated with periodontal disease
      • Lundmark A.
      • Hu Y.O.O.
      • Huss M.
      • Johannsen G.
      • Andersson A.F.
      • Yucel-Lindberg T.
      Identification of salivary microbiota and its association with host inflammatory mediators in periodontitis.
      and cystic fibrosis.
      • Flynn J.M.
      • Niccum D.
      • Dunitz J.M.
      • Hunter R.C.
      Evidence and role for bacterial mucin degradation in cystic fibrosis airway disease.
      Many bacteria that colonise the mouth express sialidases that degrade sialoglycoprotein substrates and can use sialic acid and/or underlying sugars as carbon sources, improving their survival while facilitating access to the epithelium. Sialidases are produced by oral Streptococci viridans including most strains of Streptococcus oralis, intermedius and mitis.
      • Beighton D.
      • Whiley R.A.
      Sialidase activity of the "Streptococcus milleri group" and other viridans group streptococci.
      ,
      • Byers H.L.
      • Homer K.A.
      • Beighton D.
      Utilization of sialic acid by viridans streptococci.
      Bacterial sialidases unmask underlying ligands to which bacteria or their toxins adhere.
      • Lewis A.L.
      • Lewis W.G.
      Host sialoglycans and bacterial sialidases: a mucosal perspective.
      In the saliva, rifaximin-α decreased opportunistic pathogens including Abiotrophia defectiva, Olsenella uli and Filifactor alocis, with significant increases in Lactobacillus and oral commensal species, such as Streptococcus spp, associated with oral health. Previous studies have also shown that rifaximin-α promotes the growth of beneficial strains in ulcerative colitis
      • Brigidi P.
      • Swennen E.
      • Rizzello F.
      • Bozzolasco M.
      • Matteuzzi D.
      Effects of rifaximin administration on the intestinal microbiota in patients with ulcerative colitis.
      ; additionally, in a mouse model of visceral hyperalgesia, Lactobacilli grew in the ileum in response to rifaximin-α.
      • Xu D.
      • Gao J.
      • Gillilland 3rd, M.
      • Wu X.
      • Song I.
      • Kao J.Y.
      • et al.
      Rifaximin alters intestinal bacteria and prevents stress-induced gut inflammation and visceral hyperalgesia in rats.
      Rifaximin-α reduced adhesion, invasion and motility of E. coli independent of its antimicrobial effect
      • Dogan B.
      • Fu J.
      • Zhang S.
      • Scherl E.J.
      • Simpson K.W.
      Rifaximin decreases virulence of Crohn's disease-associated Escherichia coli and epithelial inflammatory responses.
      reducing the virulence of resident microbiota.
      • Jiang Z.D.
      • Ke S.
      • Dupont H.L.
      Rifaximin-induced alteration of virulence of diarrhoea-producing Escherichia coli and Shigella sonnei.
      A recent metagenomic study evaluating patients with cirrhosis before and after rifaximin-α demonstrated collapse of bacterial–phage interactions, especially phages directed against pathobionts associated with cirrhosis, such as Streptococcus, Pseudomonas and Enterobacteriaceae spp.
      • Bajaj J.S.
      • Sikaroodi M.
      • Shamsaddini A.
      • Henseler Z.
      • Santiago-Rodriguez T.
      • Acharya C.
      • et al.
      Interaction of bacterial metagenome and virome in patients with cirrhosis and hepatic encephalopathy.
      Rifaximin-α increased faecal TNF-α and IL-17E concentrations whilst the suppression of Veillonella spp and Streptococcus spp was associated with increased faecal IL-17A. IL-17A is a mucosal-associated cytokine involved in local immune modulation. IL-17A and IL-17E are secreted by TH17 cells and play a critical role in establishing local host antimicrobial immunity and promote gut barrier repair.
      • Dixon B.R.
      • Radin J.N.
      • Piazuelo M.B.
      • Contreras D.C.
      • Algood H.M.
      IL-17a and IL-22 induce expression of antimicrobials in gastrointestinal epithelial cells and may contribute to epithelial cell defense against Helicobacter pylori.
      IL-17 and TNF-α induce antimicrobial peptides in mucosal organs with IL-17A serving as a potent neutrophil recruiter.
      • McGeachy M.J.
      • Cua D.J.
      • Gaffen S.L.
      The IL-17 family of cytokines in health and disease.
      They also promote epithelial cell proliferation and replacement of cells lost through homeostatic shedding.
      • Mendes V.
      • Galvao I.
      • Vieira A.T.
      Mechanisms by which the gut microbiota influences cytokine production and modulates host inflammatory responses.
      In addition, TNF-α is an important regulator of intestinal microbiota populations.
      • Ruder B.
      • Atreya R.
      • Becker C.
      Tumour necrosis factor Alpha in intestinal homeostasis and gut related diseases.
      IL-17E, a barrier surface cytokine, promotes epithelial cell division and increases mucus secretion.
      • Andrews C.
      • McLean M.H.
      • Durum S.K.
      Cytokine tuning of intestinal epithelial function.
      Therefore it can be postulated that rifaximin-α promotes an intestinal microenvironment conducive to increased mucus production and gut barrier repair. This may be a mechanism by which rifaximin-α reduces BT of enteropathogens and endotoxaemia although the design of this study did not allow direct interrogation of the mucus layer. Furthermore, the gut barrier is complex with a multitude of factors contributing to barrier function and immune-competence. In vitro/in vivo studies will be needed to further investigate the impact of rifaximin-α on gut barrier function which are beyond the scope of this trial.
      In summary, the mechanism of action of rifaximin-α in patients with cirrhosis has been further elucidated. Rifaximin-α improved overt HE and neurocognitive function and ameliorated systemic inflammation by suppressing oralisation of the gut microbiome via suppression of Veillonella spp and Streptococcus spp as well as Akkermansia and Hungatella; all species rich in mucin-degrading enzymes and known to induce gut barrier damage. Rifaximin-α promoted a TNF-α and IL-17E-enriched intestinal microenvironment conducive to improved antimicrobial function and gut barrier repair.

       Abbreviations

      BT, bacterial translocation; CAID, cirrhosis-associated immune dysfunction; CAZyme, carbohydrate-active enzyme; HE, hepatic encephalopathy; HRQoL, health-related quality of life; LPS, lipopolysaccharide; MGS, metagenomic species; OB, oxidative burst; PHES, psychometric hepatic encephalopathy score; PLS-DA, partial least square discriminant analysis; ROS, reactive oxygen species; TLR, Toll-like receptor; TNF-α, tumour necrosis factor-α.

      Financial support

      This trial was funded through an investigator-initiated study grant awarded by Norgine Pharmaceuticals UK Limited to King’s College London and the integrative and functional analysis was also supported by the Engineering and Physical Sciences Research Council (EPSRC) , EP/S001301/1 , and Science for Life Laboratory. The computational infrastructure to support this study was provided by the Swedish National Infrastructure for Computing at SNIC through Uppsala Multidisciplinary Centre for Advanced Computational Science (UPPMAX) under Project SNIC 2018/3-434, SNIC 2019/3-226 and SNIC 2020/6-153. Additional financial support was provided from the MetaGenoPolis grant ANR-11-DPBS-0001 . Infrastructure to support this study was also provided by the Medical Research Council (MRC) Centre for Transplantation, King's College London, UK – MRC grant no. MR/J006742/1 . This study represents independent research supported by the National Institute for Health Research (NIHR) -Wellcome King’s Clinical Research Facility and the NIHR Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London . The views expressed are those of the author (s) and not necessarily those of the NHS, NIHR, EPRSC or the Department of Health and Social Care. The bioinformatic analysis was additionally supported by the Global University Project (GUP), "GIST Research Institute (GRI) IIBR" with grants funded by the GIST in 2021, and the Bio-Synergy Research Project ( 2021M3A9C4000991 ), the Bio & Medical Technology Development Program ( 2021M3A9G8022959 ), and Basic Science Research Program ( 2021R1C1C1006336 ) of the Ministry of Science, ICT through the National Research Foundation , Korea.

      Authors’ contributions

      DLS conceived the study and served as Chief Investigator. VCP co-designed the study with DLS. VCP served as the lead investigator, recruiting the patients, undertaking data collection and coordinating the analyses with the support of AZ who recruited patients, undertook data collection and patient follow-up. MJWM performed the statistical analyses. SL and EW performed the saliva microbiome analyses. KDS and SG performed both faecal and saliva microbiome analyses. SL and SS3,8 performed the downstream analyses on the faeces and saliva microbiome data, integrative and functional analyses. SS4, GKMV, XH, SG, MC and LAE undertook the remaining laboratory analyses. ELC and NP supervised the microbiome analyses. NG generated microbiome sequence data and SDE conceived and coordinated the microbiome data generation and analyses serving as a co-investigator. JW and KB were involved throughout the trial as co-investigators. The manuscript was written by DLS and critically reviewed by all co-authors who approved the final submitted manuscript.

      Data availability statement

      All the metagenomic data are publically available in the European Nucleotide Archive: https://www.ebi.ac.uk/ena/browser/home; Identifier number: PRJEB38481. All remaining data including the metabonomic data will be made available on request. All requests should be sent to the corresponding author: [email protected] .

      Conflict of interest

      Dr V Patel and Ms A Zamalloa have delivered paid lectures for Norgine Pharmaceuticals Ltd. Professor Shawcross has participated in advisory boards for Norgine Pharmaceuticals Ltd, EnteroBiotix, Kaleido Biosciences, Mallinckrodt and Shionogi and has delivered paid lectures for Norgine Pharmaceuticals Ltd, Falk Pharma and Alfa Sigma.
      Please refer to the accompanying ICMJE disclosure forms for further details.

      Supplementary data

      References

        • Patidar K.R.
        • Thacker L.R.
        • Wade J.B.
        • Sterling R.K.
        • Sanyal A.J.
        • Siddiqui M.S.
        • et al.
        Covert hepatic encephalopathy is independently associated with poor survival and increased risk of hospitalization.
        Am J Gastroenterol. 2014; 109: 1757-1763
        • Cordoba J.
        • Ventura-Cots M.
        • Simon-Talero M.
        • Amoros A.
        • Pavesi M.
        • Vilstrup H.
        • et al.
        Characteristics, risk factors, and mortality of cirrhotic patients hospitalized for hepatic encephalopathy with and without acute-on-chronic liver failure.
        J Hepatol. 2014; 60: 275-281
        • Woodhouse C.A.
        • Patel V.C.
        • Singanayagam A.
        • Shawcross D.L.
        The gut microbiome as a therapeutic target in the pathogenesis and treatment of chronic liver disease.
        Aliment Pharmacol Ther. 2018; 47: 192-202
        • Bajaj J.S.
        • Heuman D.M.
        • Sanyal A.J.
        • Hylemon P.B.
        • Sterling R.K.
        • Stravitz R.T.
        • et al.
        Modulation of the metabiome by rifaximin in patients with cirrhosis and minimal hepatic encephalopathy.
        PLoS One. 2013; 8e60042
        • Bajaj J.S.
        • Ridlon J.M.
        • Hylemon P.B.
        • Thacker L.R.
        • Heuman D.M.
        • Smith S.
        • et al.
        Linkage of gut microbiome with cognition in hepatic encephalopathy.
        Am J Physiol Gastrointest Liver Physiol. 2012; 302: G168-G175
        • Albillos A.
        • Lario M.
        • Alvarez-Mon M.
        Cirrhosis-associated immune dysfunction: distinctive features and clinical relevance.
        J Hepatol. 2014; 61: 1385-1396
        • Bellot P.
        • Garcia-Pagan J.C.
        • Frances R.
        • Abraldes J.G.
        • Navasa M.
        • Perez-Mateo M.
        • et al.
        Bacterial DNA translocation is associated with systemic circulatory abnormalities and intrahepatic endothelial dysfunction in patients with cirrhosis.
        Hepatology. 2010; 52: 2044-2052
        • Riva A.
        • Gray E.H.
        • Azarian S.
        • Zamalloa A.
        • McPhail M.J.
        • Vincent R.P.
        • et al.
        Faecal cytokine profiling as a marker of intestinal inflammation in acutely decompensated cirrhosis.
        JHEP Rep. 2020; 2: 100151
        • Bass N.M.
        • Mullen K.D.
        • Sanyal A.
        • Poordad F.
        • Neff G.
        • Leevy C.B.
        • et al.
        Rifaximin treatment in hepatic encephalopathy.
        N Engl J Med. 2010; 362: 1071-1081
        • Orr J.G.
        • Currie C.J.
        • Berni E.
        • Goel A.
        • Moriarty K.J.
        • Sinha A.
        • et al.
        The impact on hospital resource utilisation of treatment of hepatic encephalopathy with rifaximin-alpha.
        Liver Int. 2016; 36: 1295-1303
        • Kimer N.
        • Pedersen J.S.
        • Tavenier J.
        • Christensen J.E.
        • Busk T.M.
        • Hobolth L.
        • et al.
        Rifaximin has minor effects on bacterial composition, inflammation, and bacterial translocation in cirrhosis: a randomized trial.
        J Gastroenterol Hepatol. 2018; 33: 307-314
        • Trebicka J.
        • Bork P.
        • Krag A.
        • Arumugam M.
        Utilizing the gut microbiome in decompensated cirrhosis and acute-on-chronic liver failure.
        Nat Rev Gastroenterol Hepatol. 2021; 18: 167-180
        • Conn H.O.
        • Leevy C.M.
        • Vlahcevic Z.R.
        • Rodgers J.B.
        • Maddrey W.C.
        • Seeff L.
        • et al.
        Comparison of lactulose and neomycin in the treatment of chronic portal-systemic encephalopathy. A double blind controlled trial.
        Gastroenterology. 1977; 72: 573-583
        • Weissenborn K.
        • Ennen J.C.
        • Schomerus H.
        • Ruckert N.
        • Hecker H.
        Neuropsychological characterization of hepatic encephalopathy.
        J Hepatol. 2001; 34: 768-773
        • Dolan P.
        Modeling valuations for EuroQol health states.
        Med Care. 1997; 35: 1095-1108
        • Taylor N.J.
        • Manakkat Vijay G.K.
        • Abeles R.D.
        • Auzinger G.
        • Bernal W.
        • Ma Y.
        • et al.
        The severity of circulating neutrophil dysfunction in patients with cirrhosis is associated with 90-day and 1-year mortality.
        Aliment Pharmacol Ther. 2014; 40: 705-715
        • Burri E.
        • Manz M.
        • Rothen C.
        • Rossi L.
        • Beglinger C.
        • Lehmann F.S.
        Monoclonal antibody testing for fecal calprotectin is superior to polyclonal testing of fecal calprotectin and lactoferrin to identify organic intestinal disease in patients with abdominal discomfort.
        Clin Chim Acta. 2013; 416: 41-47
        • McPhail M.J.W.
        • Shawcross D.L.
        • Lewis M.R.
        • Coltart I.
        • Want E.J.
        • Antoniades C.G.
        • et al.
        Multivariate metabotyping of plasma predicts survival in patients with decompensated cirrhosis.
        J Hepatol. 2016; 64: 1058-1067
        • Dona A.C.
        • Jimenez B.
        • Schafer H.
        • Humpfer E.
        • Spraul M.
        • Lewis M.R.
        • et al.
        Precision high-throughput proton NMR spectroscopy of human urine, serum, and plasma for large-scale metabolic phenotyping.
        Anal Chem. 2014; 86: 9887-9894
        • Beckonert O.
        • Keun H.C.
        • Ebbels T.M.
        • Bundy J.
        • Holmes E.
        • Lindon J.C.
        • et al.
        Metabolic profiling, metabolomic and metabonomic procedures for NMR spectroscopy of urine, plasma, serum and tissue extracts.
        Nat Protoc. 2007; 2: 2692-2703
        • Lelouvier B.
        • Servant F.
        • Paisse S.
        • Brunet A.C.
        • Benyahya S.
        • Serino M.
        • et al.
        Changes in blood microbiota profiles associated with liver fibrosis in obese patients: a pilot analysis.
        Hepatology. 2016; 64: 2015-2027
        • Edwards L.A.
        • Nistala K.
        • Mills D.C.
        • Stephenson H.N.
        • Zilbauer M.
        • Wren B.W.
        • et al.
        Delineation of the innate and adaptive T-cell immune outcome in the human host in response to Campylobacter jejuni infection.
        PLoS One. 2010; 5e15398
        • Human Microbiome Project C
        Structure, function and diversity of the healthy human microbiome.
        Nature. 2012; 486: 207-214
        • Salehi S.
        • Tranah T.H.
        • Lim S.
        • Heaton N.
        • Heneghan M.
        • Aluvihare V.
        • et al.
        Rifaximin reduces the incidence of spontaneous bacterial peritonitis, variceal bleeding and all-cause admissions in patients on the liver transplant waiting list.
        Aliment Pharmacol Ther. 2019; 50: 435-441
        • Menshawy A.
        • Mattar O.
        • Barssoum K.
        • AboEl-Naga A.M.
        • Salim H.M.
        • Mohamed A.M.F.
        • et al.
        Safety and efficacy of rifaximin in prophylaxis of spontaneous bacterial peritonitis: a systematic review and meta-analysis.
        Curr Drug Targets. 2019; 20: 380-387
        • Bajaj J.S.
        • Heuman D.M.
        • Hylemon P.B.
        • Sanyal A.J.
        • White M.B.
        • Monteith P.
        • et al.
        Altered profile of human gut microbiome is associated with cirrhosis and its complications.
        J Hepatol. 2014; 60: 940-947
        • Qin N.
        • Yang F.
        • Li A.
        • Prifti E.
        • Chen Y.
        • Shao L.
        • et al.
        Alterations of the human gut microbiome in liver cirrhosis.
        Nature. 2014; 513: 59-64
        • Bajaj J.S.
        • Betrapally N.S.
        • Hylemon P.B.
        • Heuman D.M.
        • Daita K.
        • White M.B.
        • et al.
        Salivary microbiota reflects changes in gut microbiota in cirrhosis with hepatic encephalopathy.
        Hepatology. 2015; 62: 1260-1271
        • Chalmers N.I.
        • Palmer Jr., R.J.
        • Cisar J.O.
        • Kolenbrander P.E.
        Characterization of a Streptococcus sp.-Veillonella sp. community micromanipulated from dental plaque.
        J Bacteriol. 2008; 190: 8145-8154
        • Lundmark A.
        • Hu Y.O.O.
        • Huss M.
        • Johannsen G.
        • Andersson A.F.
        • Yucel-Lindberg T.
        Identification of salivary microbiota and its association with host inflammatory mediators in periodontitis.
        Front Cell Infect Microbiol. 2019; 9: 216
        • Flynn J.M.
        • Niccum D.
        • Dunitz J.M.
        • Hunter R.C.
        Evidence and role for bacterial mucin degradation in cystic fibrosis airway disease.
        PLoS Pathog. 2016; 12e1005846
        • Beighton D.
        • Whiley R.A.
        Sialidase activity of the "Streptococcus milleri group" and other viridans group streptococci.
        J Clin Microbiol. 1990; 28: 1431-1433
        • Byers H.L.
        • Homer K.A.
        • Beighton D.
        Utilization of sialic acid by viridans streptococci.
        J Dent Res. 1996; 75: 1564-1571
        • Lewis A.L.
        • Lewis W.G.
        Host sialoglycans and bacterial sialidases: a mucosal perspective.
        Cell Microbiol. 2012; 14: 1174-1182
        • Brigidi P.
        • Swennen E.
        • Rizzello F.
        • Bozzolasco M.
        • Matteuzzi D.
        Effects of rifaximin administration on the intestinal microbiota in patients with ulcerative colitis.
        J Chemother. 2002; 14: 290-295
        • Xu D.
        • Gao J.
        • Gillilland 3rd, M.
        • Wu X.
        • Song I.
        • Kao J.Y.
        • et al.
        Rifaximin alters intestinal bacteria and prevents stress-induced gut inflammation and visceral hyperalgesia in rats.
        Gastroenterology. 2014; 146: 484-496 e4
        • Dogan B.
        • Fu J.
        • Zhang S.
        • Scherl E.J.
        • Simpson K.W.
        Rifaximin decreases virulence of Crohn's disease-associated Escherichia coli and epithelial inflammatory responses.
        J Antibiot (Tokyo). 2018; 71: 485-494
        • Jiang Z.D.
        • Ke S.
        • Dupont H.L.
        Rifaximin-induced alteration of virulence of diarrhoea-producing Escherichia coli and Shigella sonnei.
        Int J Antimicrob Agents. 2010; 35: 278-281
        • Bajaj J.S.
        • Sikaroodi M.
        • Shamsaddini A.
        • Henseler Z.
        • Santiago-Rodriguez T.
        • Acharya C.
        • et al.
        Interaction of bacterial metagenome and virome in patients with cirrhosis and hepatic encephalopathy.
        Gut. 2021; 70: 1162-1173
        • Dixon B.R.
        • Radin J.N.
        • Piazuelo M.B.
        • Contreras D.C.
        • Algood H.M.
        IL-17a and IL-22 induce expression of antimicrobials in gastrointestinal epithelial cells and may contribute to epithelial cell defense against Helicobacter pylori.
        PLoS One. 2016; 11e0148514
        • McGeachy M.J.
        • Cua D.J.
        • Gaffen S.L.
        The IL-17 family of cytokines in health and disease.
        Immunity. 2019; 50: 892-906
        • Mendes V.
        • Galvao I.
        • Vieira A.T.
        Mechanisms by which the gut microbiota influences cytokine production and modulates host inflammatory responses.
        J Interferon Cytokine Res. 2019; 39: 393-409
        • Ruder B.
        • Atreya R.
        • Becker C.
        Tumour necrosis factor Alpha in intestinal homeostasis and gut related diseases.
        Int J Mol Sci. 2019; 20
        • Andrews C.
        • McLean M.H.
        • Durum S.K.
        Cytokine tuning of intestinal epithelial function.
        Front Immunol. 2018; 9: 1270